This page provides RoB assessments and the rationale for those judgments based on the excerpts in the examples sections of each of the domain pages. Assessments can be:
LOW risk of bias
UNCLEAR risk of bias
HIGH risk of bias
Please leave a comment if you interpreted things differently!
1. Children were assigned to receive vaccine or placebo according to a schedule for randomly assigning numbers that was based on the sequential order of injection.
Judgment:UNCLEAR
Rationale: Insufficient data is provided to determine whether the sequence generation method is actually random.
Citation: Werzberger A, Mensch B, Kuter B, Brown L, Lewis J, Sitrin R, Miller W, Shouval D, Wiens B, Calandra G, Ryan J, Provost P, Nalin D. A controlled trial of a formalin-inactivated hepatitis A vaccine in healthy children. New England Journal of Medicine 1992;327:453-457. 2. The randomization sequence was computer-generated and known only to one author who played no part in the enrolment of participants.
Judgment: LOW
Rationale: Description of an adequate random component in the sequence generation process.
Citation: Zar HJ, Streun S, Levin M, Weinberg EG, Swingler GH. Randomised controlled trial of the efficacy of a metered dose inhaler with bottle spacer for bronchodilator treatment in acute lower airway obstruction. Archives of Disease in Childhood 2007;92:142-146. 3. Infants were assigned to intervention or control by odd or even hospital case number.
Judgment: HIGH
Rationale: Hospital case numbers are not assigned randomly, therefore this is an inappropriate method of sequence generation.
Citation: Cruz H, Guzman N, Rosales M, Bastidas J, Garcia J, Hurtado I, Rojas ME, Hodgman JE. Early hospital discharge of preterm very low birth weight infants. Journal of Perinatology 1997;17:29-32. 4. Infants receiving opiates were randomly assigned by medical record number to receive either fentanyl citrate or morphine sulfate as single intravenous doses.
Judgment: UNCLEAR / HIGH
Rationale (Unclear): Medical record numbers may have been appropriately randomized, but there is insufficient data to make a judgment. Rationale (High): Assignment based on medical record numbers is an inappropriate method of sequence generation.
Citation: Ionides SP, Weiss MG, Angelopoulos M, Myers TF, Handa RJ. Plasma beta-endorphin concentrations and analgesia-muscle relaxation in the newborn infant supported by mechanical ventilation. Journal of Pediatrics 1994;125:113-116. 5. Treatment allocation was made on a temporal basis at two sites. Assignment to a particular study group was dependent on the site, with monthly rotation of dosing regimens.
Judgment: UNCLEAR
Rationale: Unclear on how the sequence was generated and whether or not it was truly random.
Citation: Thureen PJ, Reiter PD, Gresores A, Stolpman NM, Kawato K, Hall DM. Once- versus twice-daily gentamicin dosing in neonates ≥34 weeks’ gestation: cost-effectiveness analysis. Pediatrics 1999;103:594-598. 6. Before study initiation, a list of randomization numbers and corresponding treatment numbers was computer generated by a third party.
Judgment: LOW
Rationale: Description of an adequate random component in the sequence generation process.
Citation: Di Prospero NA, Baker A, Jeffries N, Fischbeck KH. Neurological effects of high-dose idebenone in patients with Friedrich’s ataxia: a randomised, placebo-controlled trial. Lancet Neurology 2007;6:878-886.
7. Patients were randomly selected using a block randomization procedure.
Judgment: UNCLEAR
Rationale: A description of block randomization alone is not sufficient for a low risk rating unless accompanied by a description indicating the sequence was computer generated.
Citation: Ment LR, Oh W, Ehrenkranz RA, Philip AG, Vohr B, Allan W, Duncan CC, Scott DT, Taylor KJ, Katz KH, Schneider KC, Makuch RW. Low-dose indomethacin and prevention of intraventricular hemorrhage: a multicenter randomized trial. Pediatrics 1994;93:543-550. [back to top] [Sequence generation]
Allocation concealment
1. Randomization was done by shuffled, color-coded cards that were assigned once the infant became eligible for treatment. The intervention was coded as pink and the control was coded as green. The cards were shuffled only once. Before each treatment, the attending neonatologist picked a colored card at the top of the deck and the corresponding intervention was administered.
Judgment: HIGH
Rationale: Color-coding the cards would allow the investigators to foresee the group assignment.
Citation: Modanlou HD, Beharry K, Padilla G, Norris K, Safvati S, Aranda JV. Comparative efficacy of exosurf and survanta surfactants on early clinical course of respiratory distress syndrome and complications of prematurity. Journal of Perinatology 1997;17:455-460.
2. Allocation was concealed until after the intervention had been assigned, using sequentially numbered sealed envelopes that contained tin foil to make them opaque to bright light.
Judgment: LOW
Rationale: Measures were taken to ensure that the group assignment would not be apparent to the investigators prior to allocation.
Citation: Zar HJ, Streun S, Levin M, Weinberg EG, Swingler GH. Randomised controlled trial of the efficacy of a metered dose inhaler with bottle spacer for bronchodilator treatment in acute lower airway obstruction. Archives of Disease in Childhood 2007;92:142-146.
3. The schedules for the group assignments were in 3-mo blocks for the first 2 y and 2-mo blocks for the third year. The first 3-mo block was randomized, after which each block was alternated to intervention or control. The order of alternation was reversed after the first year. Due to the relatively large population involved, the statistician deemed this method of assignment appropriate.
Judgment: HIGH
Rationale: Alternation would allow the investigators to be aware of the group assignments, or have a high likelihood of being able to discern the group assignments given the two or three month period of assignment.
Citation: Patel DM, Rhodes PG, LeBlanc MH, Graves GR, Glick C, Morrison J. Role of postnatal penicillin prophylaxis in prevention of neonatal group B streptococcus infection. Acta Paediatrica 1999;88:874-879.
4. The treatment assignments were maintained by the third party and only made available after the trial was complete and the database finalized.
Judgment: LOW
Rationale: Randomization sequence maintained by a third party uninvolved in the investigation.
Citation: Di Prospero NA, Baker A, Jeffries N, Fischbeck KH. Neurological effects of high-dose idebenone in patients with Friedrich’s ataxia: a randomised, placebo-controlled trial. Lancet Neurology 2007;6:878-886.
[back to top] [Allocation concealment]
Blinding of participants and study personnel
1. In order to conceal the allocation sequence, the pharmacy at each site prepared the study drugs in sequentially numbered, visually identical packets. The active drugs and placebo were identical in appearance, volume, weight, odor and taste.
Judgment: LOW
Rationale: Blinding of participants and key study personnel ensured, and unlikely that it could have been broken.
Citation: Plint AC, Johnson DW, Patel H, Wiebe N, Correll R, Brant R, Mitton C, Goudin S, Bhatt M, Joubert G, Black KJL, Turner T, Whitehouse S, Klassen TP for PERC. Epinephrine and dexamethasone in children with bronchiolitis. New England Journal of Medicine 2009;360:2079-2089.
2. Patients were randomly assigned to receive oral salbutamol or MDI salbutamol or formed the control group (without any bronchodilator therapy).
Judgment: HIGH
Rationale: There is no blinding with an oral treatment, an inhaled treatment, and no treatment.
Citation: Cengizlier R, Saraclar Y, Adalioglu G, Tuncer A. Effect of oral and inhaled salbutamol in infants with bronchiolitis. Acta Paediatrica Japonica; Overseas edition 1997;39:61-63.
3. Questions were raised midway through the trial about whether there was an increased incidence of necrotizing enterocolitis in the study population. The DSMC noted no significant difference in incidence of necrotizing enterocolitis between groups. There were, however, fewer nosocomial infections among the immune globulin recipients; hence, the committee recommended continuing the study, but since the placebo infusion could not be beneficial and could instead be harmful, the committee mandated that the placebo infusion be discontinued and that the study continue in an unblinded fashion (immune globulin or no infusion).
Judgment: HIGH
Rationale: There is no blinding with a treatment being compared to no treatment, potentially influencing personnel behaviour.
Citation: Fanaroff AA, Korones SB, Wright LL, Wright EC, Poland RL, Bauer CB, Tyson JE, Philips JB, Edwards W, Lucey JF, Catz CS, Shankaran S, Oh W for the National Institute of Child Health and Human Development Neonatal Research Network. A controlled trial of intravenous immune globulin to reduce nosocomial infections in very-low-birth-weight infants. New England Journal of Medicine 1994;330:1107-1113.
4. This is a non-blinded, randomised prospective comparative study. Randomisation to two separate groups was done according to file number at time of admission. Two physicians, who did not know which treatment was applied to the infant, performed the clinical assessment, and clinical score was calculated as average of the two measurements.
Judgment: UNCLEAR
Rationale: The study is described as non-blinded, but two blinded physicians performed the outcome assessment. It is unclear whether the outcome assessors are separate from the personnel that would be administering the treatment.
Citation: Uyan AP, Ozyurek H, Keskin M, Afsar Y, Yilmaz E. Comparison of two different bronchodilators in the treatment of acute bronchiolitis. The Internet Journal of Pediatrics and Neonatology 2003;3.
[back to top] [Blinding]
Blinding of outcome assessors
1. This is a non-blinded, randomised prospective comparative study. Randomisation to two separate groups was done according to file number at time of admission. Two physicians, who did not know which treatment was applied to the infant, performed the clinical assessment, and clinical score was calculated as average of the two measurements.
Judgment: LOW
Rationale: The outcome assessors are explicitly described as being blinded.
Citation: Uyan AP, Ozyurek H, Keskin M, Afsar Y, Yilmaz E. Comparison of two different bronchodilators in the treatment of acute bronchiolitis. The Internet Journal of Pediatrics and Neonatology 2003;3.
2. Randomization occurred in two steps. Children were first assigned to either a medication or surgical group. Those in the medication group were then randomized again at the Pharmacy Department into either a placebo group or a sulfisoxazole group.
Judgment: HIGH
Rationale: It would be apparent to the individual assessing for otitis media whether the participant had been assigned to the surgical or medication group.
Citation: Gonzalez C, Arnold JE, Woody EA, Erhardt JB, Pratt SR, Getts A, Kueser TJ, Kolmer JW, Sachs M. Prevention of recurrent acute otitis media: chemoprophylaxis versus tympanostomy tubes. Laryngoscope 1986;96:1330-1334.
[back to top] [Blinding]
Incomplete outcome data
1. 102 were considered eligible for inclusion…13 children were subsequently excluded because 12 of them developed pulmonary consolidation and were put on antibiotics by the treating physician and in one deterioration of the clinical status required transfer to the ICU. A total of 89 patients therefore completed the study.
Judgment: UNCLEAR
Rationale: Excluded participants are described, but the number from each group is not mentioned - it is unclear whether the withdrawals are balanced across treatment conditions.
Citation: Chowdhury D, al Howasi M, Khalil M, al-Frayh AS, Chowdhury S, Ramia S. The role of bronchodilators in the management of bronchiolitis: a clinical trial. Annals of Tropical Paediatrics 1995;15:77-84.
2. Thirty-one infants were enrolled in the study.
Judgment: UNCLEAR
Rationale: No details are provided on how many completed the study.
Citation: Cengizlier R, Saraclar Y, Adalioglu G, Tuncer A. Effect of oral and inhaled salbutamol in infants with bronchiolitis. Acta Paediatrica Japonica; Overseas edition 1997;39:61-63.
3. There were no protocol deviations and no drug discontinuations, although one patient in the low-dose group could not undergo follow-up testing because of intercurrent illness and was excluded for the final analysis of ICARS and FARS (these data were not imputed), but was included in the primary endpoint analysis and ADL assessment.
Judgment: LOW
Rationale: All participants are described and accounted for, and all are included in the primary analysis.
Citation: Di Prospero NA, Baker A, Jeffries N, Fischbeck KH. Neurological effects of high-dose idebenone in patients with Friedrich’s ataxia: a randomised, placebo-controlled trial. Lancet Neurology 2007;6:878-886.
4. Of the 88 children completing the study, 13 infants were included in a pilot, in which they received only one treatment and were assessed at 30 minutes. Seventy-six infants completed evaluations 30 and 60 minutes after the first treatment.
Judgment: UNCLEAR
Rationale: It is unclear how many children were enrolled (and also whether they included an extra participant in their description, or if there is a typo).
Citation: Gadomski AM, Lichenstein R, Horton L, King J, Keane V, Permutt T. Efficacy of albuterol in the management of bronchiolitis. Pediatrics 1994;93:907-912.
5. Thirty-two patients fulfilled all entry criteria. Three patients did not complete the study and were not included for analysis.
Judgment: UNCLEAR
Rationale: Three patients out of thirty-two is 9%. If these were all from the same study arm, bias may have been introduced.
Citation: De Boeck K, Van der Aa N, Van Lierde S, Corbeel L, Eeckels R. Respiratory syncytial virus bronchiolitis: A double-blind dexamethasone efficacy study. Journal of Pediatrics 1997;131:919-921.
[back to top] [Incomplete outcome data]
Selective outcome reporting
1. In data analysis section: No other comparisons were planned or performed.
Judgment: LOW
Rationale: There is a clear statement that the analyses that were outlined in the Methods section match the trial design.
Citation: Zar HJ, Streun S, Levin M, Weinberg EG, Swingler GH. Randomised controlled trial of the efficacy of a metered dose inhaler with bottle spacer for bronchodilator treatment in acute lower airway obstruction. Archives of Disease in Childhood 2007;92:142-146.
2. Post-hoc analyses demonstrated that stride length was the gait parameter largely contributing to the overall difference in general gait patterns.
Judgment: UNCLEAR
Rationale: It is unclear why this post-hoc analysis was performed.
Citation: Wu J, Looper J, Ulrich BD, Ulrich DA, Angulo-Barroso RM. Exploring effects of different treadmill interventions on walking onset and gait patterns in infants with Down Syndrome. Developmental Medicine and Child Neurology 2007;49:839-845.
[back to top] [Selective outcome reporting]
Other sources of bias
1. The demographic and health status characteristics in the three groups did not differ significantly with respect to age, sex, passive smoking, family history of recurrent AOM, history of RAD, use of child care, age of initial otitis episode, history of supine feeding, mouth breathing or snoring (Table 1). The frequency of upper respiratory infections experienced by subjects during the study was also similar for the three groups.
Judgment: LOW
Rationale: Baseline characteristics are balanced.
Citation: Roark R, Berman S. Continuous twice daily or once daily amoxicillin prophylaxis compared with placebo for children with recurrent acute otitis media. Pediatric Infectious Disease Journal 1997;16:376-381.
2. Randomization was accomplished using a randomized block design in which block size was randomly allocated between 2 and 4 to ensure that the size of the intervention and control groups was equivalent. […] Children randomized to the control group received routine care […]; children who were randomized to the intervention group received an asthma education and management intervention.
Judgment: HIGH
Rationale: Given that this was an educational intervention, this was an unblinded trial. Using blocks of 2-4 would allow for easy determination of the group allocation by the investigators prior to assignment.
Citation: Farber HJ, Oliveria L. Trial of an asthma education program in an inner-city pediatric emergency department. Pediatric Asthma, Allergy & Immunology 2004;17:107-115.
[back to top] [Other sources of bias]
Please leave a comment if you interpreted things differently!
Table of Contents
To leave a comment, use the 'Discussion Posts' icon in the grey bar on the top right-hand side of your screen.
Risk of Bias Guidelines
Download examples:
Sequence generation
1. Children were assigned to receive vaccine or placebo according to a schedule for randomly assigning numbers that was based on the sequential order of injection.
Judgment: UNCLEAR
Rationale: Insufficient data is provided to determine whether the sequence generation method is actually random.
Citation: Werzberger A, Mensch B, Kuter B, Brown L, Lewis J, Sitrin R, Miller W, Shouval D, Wiens B, Calandra G, Ryan J, Provost P, Nalin D. A controlled trial of a formalin-inactivated hepatitis A vaccine in healthy children. New England Journal of Medicine 1992;327:453-457.
2. The randomization sequence was computer-generated and known only to one author who played no part in the enrolment of participants.
Judgment: LOW
Rationale: Description of an adequate random component in the sequence generation process.
Citation: Zar HJ, Streun S, Levin M, Weinberg EG, Swingler GH. Randomised controlled trial of the efficacy of a metered dose inhaler with bottle spacer for bronchodilator treatment in acute lower airway obstruction. Archives of Disease in Childhood 2007;92:142-146.
3. Infants were assigned to intervention or control by odd or even hospital case number.
Judgment: HIGH
Rationale: Hospital case numbers are not assigned randomly, therefore this is an inappropriate method of sequence generation.
Citation: Cruz H, Guzman N, Rosales M, Bastidas J, Garcia J, Hurtado I, Rojas ME, Hodgman JE. Early hospital discharge of preterm very low birth weight infants. Journal of Perinatology 1997;17:29-32.
4. Infants receiving opiates were randomly assigned by medical record number to receive either fentanyl citrate or morphine sulfate as single intravenous doses.
Judgment: UNCLEAR / HIGH
Rationale (Unclear): Medical record numbers may have been appropriately randomized, but there is insufficient data to make a judgment.
Rationale (High): Assignment based on medical record numbers is an inappropriate method of sequence generation.
Citation: Ionides SP, Weiss MG, Angelopoulos M, Myers TF, Handa RJ. Plasma beta-endorphin concentrations and analgesia-muscle relaxation in the newborn infant supported by mechanical ventilation. Journal of Pediatrics 1994;125:113-116.
5. Treatment allocation was made on a temporal basis at two sites. Assignment to a particular study group was dependent on the site, with monthly rotation of dosing regimens.
Judgment: UNCLEAR
Rationale: Unclear on how the sequence was generated and whether or not it was truly random.
Citation: Thureen PJ, Reiter PD, Gresores A, Stolpman NM, Kawato K, Hall DM. Once- versus twice-daily gentamicin dosing in neonates ≥34 weeks’ gestation: cost-effectiveness analysis. Pediatrics 1999;103:594-598.
6. Before study initiation, a list of randomization numbers and corresponding treatment numbers was computer generated by a third party.
Judgment: LOW
Rationale: Description of an adequate random component in the sequence generation process.
Citation: Di Prospero NA, Baker A, Jeffries N, Fischbeck KH. Neurological effects of high-dose idebenone in patients with Friedrich’s ataxia: a randomised, placebo-controlled trial. Lancet Neurology 2007;6:878-886.
7. Patients were randomly selected using a block randomization procedure.
Judgment: UNCLEAR
Rationale: A description of block randomization alone is not sufficient for a low risk rating unless accompanied by a description indicating the sequence was computer generated.
Citation: Ment LR, Oh W, Ehrenkranz RA, Philip AG, Vohr B, Allan W, Duncan CC, Scott DT, Taylor KJ, Katz KH, Schneider KC, Makuch RW. Low-dose indomethacin and prevention of intraventricular hemorrhage: a multicenter randomized trial. Pediatrics 1994;93:543-550.
[back to top] [Sequence generation]
Allocation concealment
1. Randomization was done by shuffled, color-coded cards that were assigned once the infant became eligible for treatment. The intervention was coded as pink and the control was coded as green. The cards were shuffled only once. Before each treatment, the attending neonatologist picked a colored card at the top of the deck and the corresponding intervention was administered.
Judgment: HIGH
Rationale: Color-coding the cards would allow the investigators to foresee the group assignment.
Citation: Modanlou HD, Beharry K, Padilla G, Norris K, Safvati S, Aranda JV. Comparative efficacy of exosurf and survanta surfactants on early clinical course of respiratory distress syndrome and complications of prematurity. Journal of Perinatology 1997;17:455-460.
2. Allocation was concealed until after the intervention had been assigned, using sequentially numbered sealed envelopes that contained tin foil to make them opaque to bright light.
Judgment: LOW
Rationale: Measures were taken to ensure that the group assignment would not be apparent to the investigators prior to allocation.
Citation: Zar HJ, Streun S, Levin M, Weinberg EG, Swingler GH. Randomised controlled trial of the efficacy of a metered dose inhaler with bottle spacer for bronchodilator treatment in acute lower airway obstruction. Archives of Disease in Childhood 2007;92:142-146.
3. The schedules for the group assignments were in 3-mo blocks for the first 2 y and 2-mo blocks for the third year. The first 3-mo block was randomized, after which each block was alternated to intervention or control. The order of alternation was reversed after the first year. Due to the relatively large population involved, the statistician deemed this method of assignment appropriate.
Judgment: HIGH
Rationale: Alternation would allow the investigators to be aware of the group assignments, or have a high likelihood of being able to discern the group assignments given the two or three month period of assignment.
Citation: Patel DM, Rhodes PG, LeBlanc MH, Graves GR, Glick C, Morrison J. Role of postnatal penicillin prophylaxis in prevention of neonatal group B streptococcus infection. Acta Paediatrica 1999;88:874-879.
4. The treatment assignments were maintained by the third party and only made available after the trial was complete and the database finalized.
Judgment: LOW
Rationale: Randomization sequence maintained by a third party uninvolved in the investigation.
Citation: Di Prospero NA, Baker A, Jeffries N, Fischbeck KH. Neurological effects of high-dose idebenone in patients with Friedrich’s ataxia: a randomised, placebo-controlled trial. Lancet Neurology 2007;6:878-886.
[back to top] [Allocation concealment]
Blinding of participants and study personnel
1. In order to conceal the allocation sequence, the pharmacy at each site prepared the study drugs in sequentially numbered, visually identical packets. The active drugs and placebo were identical in appearance, volume, weight, odor and taste.
Judgment: LOW
Rationale: Blinding of participants and key study personnel ensured, and unlikely that it could have been broken.
Citation: Plint AC, Johnson DW, Patel H, Wiebe N, Correll R, Brant R, Mitton C, Goudin S, Bhatt M, Joubert G, Black KJL, Turner T, Whitehouse S, Klassen TP for PERC. Epinephrine and dexamethasone in children with bronchiolitis. New England Journal of Medicine 2009;360:2079-2089.
2. Patients were randomly assigned to receive oral salbutamol or MDI salbutamol or formed the control group (without any bronchodilator therapy).
Judgment: HIGH
Rationale: There is no blinding with an oral treatment, an inhaled treatment, and no treatment.
Citation: Cengizlier R, Saraclar Y, Adalioglu G, Tuncer A. Effect of oral and inhaled salbutamol in infants with bronchiolitis. Acta Paediatrica Japonica; Overseas edition 1997;39:61-63.
3. Questions were raised midway through the trial about whether there was an increased incidence of necrotizing enterocolitis in the study population. The DSMC noted no significant difference in incidence of necrotizing enterocolitis between groups. There were, however, fewer nosocomial infections among the immune globulin recipients; hence, the committee recommended continuing the study, but since the placebo infusion could not be beneficial and could instead be harmful, the committee mandated that the placebo infusion be discontinued and that the study continue in an unblinded fashion (immune globulin or no infusion).
Judgment: HIGH
Rationale: There is no blinding with a treatment being compared to no treatment, potentially influencing personnel behaviour.
Citation: Fanaroff AA, Korones SB, Wright LL, Wright EC, Poland RL, Bauer CB, Tyson JE, Philips JB, Edwards W, Lucey JF, Catz CS, Shankaran S, Oh W for the National Institute of Child Health and Human Development Neonatal Research Network. A controlled trial of intravenous immune globulin to reduce nosocomial infections in very-low-birth-weight infants. New England Journal of Medicine 1994;330:1107-1113.
4. This is a non-blinded, randomised prospective comparative study. Randomisation to two separate groups was done according to file number at time of admission. Two physicians, who did not know which treatment was applied to the infant, performed the clinical assessment, and clinical score was calculated as average of the two measurements.
Judgment: UNCLEAR
Rationale: The study is described as non-blinded, but two blinded physicians performed the outcome assessment. It is unclear whether the outcome assessors are separate from the personnel that would be administering the treatment.
Citation: Uyan AP, Ozyurek H, Keskin M, Afsar Y, Yilmaz E. Comparison of two different bronchodilators in the treatment of acute bronchiolitis. The Internet Journal of Pediatrics and Neonatology 2003;3.
[back to top] [Blinding]
Blinding of outcome assessors
1. This is a non-blinded, randomised prospective comparative study. Randomisation to two separate groups was done according to file number at time of admission. Two physicians, who did not know which treatment was applied to the infant, performed the clinical assessment, and clinical score was calculated as average of the two measurements.
Judgment: LOW
Rationale: The outcome assessors are explicitly described as being blinded.
Citation: Uyan AP, Ozyurek H, Keskin M, Afsar Y, Yilmaz E. Comparison of two different bronchodilators in the treatment of acute bronchiolitis. The Internet Journal of Pediatrics and Neonatology 2003;3.
2. Randomization occurred in two steps. Children were first assigned to either a medication or surgical group. Those in the medication group were then randomized again at the Pharmacy Department into either a placebo group or a sulfisoxazole group.
Judgment: HIGH
Rationale: It would be apparent to the individual assessing for otitis media whether the participant had been assigned to the surgical or medication group.
Citation: Gonzalez C, Arnold JE, Woody EA, Erhardt JB, Pratt SR, Getts A, Kueser TJ, Kolmer JW, Sachs M. Prevention of recurrent acute otitis media: chemoprophylaxis versus tympanostomy tubes. Laryngoscope 1986;96:1330-1334.
[back to top] [Blinding]
Incomplete outcome data
1. 102 were considered eligible for inclusion…13 children were subsequently excluded because 12 of them developed pulmonary consolidation and were put on antibiotics by the treating physician and in one deterioration of the clinical status required transfer to the ICU. A total of 89 patients therefore completed the study.
Judgment: UNCLEAR
Rationale: Excluded participants are described, but the number from each group is not mentioned - it is unclear whether the withdrawals are balanced across treatment conditions.
Citation: Chowdhury D, al Howasi M, Khalil M, al-Frayh AS, Chowdhury S, Ramia S. The role of bronchodilators in the management of bronchiolitis: a clinical trial. Annals of Tropical Paediatrics 1995;15:77-84.
2. Thirty-one infants were enrolled in the study.
Judgment: UNCLEAR
Rationale: No details are provided on how many completed the study.
Citation: Cengizlier R, Saraclar Y, Adalioglu G, Tuncer A. Effect of oral and inhaled salbutamol in infants with bronchiolitis. Acta Paediatrica Japonica; Overseas edition 1997;39:61-63.
3. There were no protocol deviations and no drug discontinuations, although one patient in the low-dose group could not undergo follow-up testing because of intercurrent illness and was excluded for the final analysis of ICARS and FARS (these data were not imputed), but was included in the primary endpoint analysis and ADL assessment.
Judgment: LOW
Rationale: All participants are described and accounted for, and all are included in the primary analysis.
Citation: Di Prospero NA, Baker A, Jeffries N, Fischbeck KH. Neurological effects of high-dose idebenone in patients with Friedrich’s ataxia: a randomised, placebo-controlled trial. Lancet Neurology 2007;6:878-886.
4. Of the 88 children completing the study, 13 infants were included in a pilot, in which they received only one treatment and were assessed at 30 minutes. Seventy-six infants completed evaluations 30 and 60 minutes after the first treatment.
Judgment: UNCLEAR
Rationale: It is unclear how many children were enrolled (and also whether they included an extra participant in their description, or if there is a typo).
Citation: Gadomski AM, Lichenstein R, Horton L, King J, Keane V, Permutt T. Efficacy of albuterol in the management of bronchiolitis. Pediatrics 1994;93:907-912.
5. Thirty-two patients fulfilled all entry criteria. Three patients did not complete the study and were not included for analysis.
Judgment: UNCLEAR
Rationale: Three patients out of thirty-two is 9%. If these were all from the same study arm, bias may have been introduced.
Citation: De Boeck K, Van der Aa N, Van Lierde S, Corbeel L, Eeckels R. Respiratory syncytial virus bronchiolitis: A double-blind dexamethasone efficacy study. Journal of Pediatrics 1997;131:919-921.
[back to top] [Incomplete outcome data]
Selective outcome reporting
1. In data analysis section: No other comparisons were planned or performed.
Judgment: LOW
Rationale: There is a clear statement that the analyses that were outlined in the Methods section match the trial design.
Citation: Zar HJ, Streun S, Levin M, Weinberg EG, Swingler GH. Randomised controlled trial of the efficacy of a metered dose inhaler with bottle spacer for bronchodilator treatment in acute lower airway obstruction. Archives of Disease in Childhood 2007;92:142-146.
2. Post-hoc analyses demonstrated that stride length was the gait parameter largely contributing to the overall difference in general gait patterns.
Judgment: UNCLEAR
Rationale: It is unclear why this post-hoc analysis was performed.
Citation: Wu J, Looper J, Ulrich BD, Ulrich DA, Angulo-Barroso RM. Exploring effects of different treadmill interventions on walking onset and gait patterns in infants with Down Syndrome. Developmental Medicine and Child Neurology 2007;49:839-845.
[back to top] [Selective outcome reporting]
Other sources of bias
1. The demographic and health status characteristics in the three groups did not differ significantly with respect to age, sex, passive smoking, family history of recurrent AOM, history of RAD, use of child care, age of initial otitis episode, history of supine feeding, mouth breathing or snoring (Table 1). The frequency of upper respiratory infections experienced by subjects during the study was also similar for the three groups.
Judgment: LOW
Rationale: Baseline characteristics are balanced.
Citation: Roark R, Berman S. Continuous twice daily or once daily amoxicillin prophylaxis compared with placebo for children with recurrent acute otitis media. Pediatric Infectious Disease Journal 1997;16:376-381.
2. Randomization was accomplished using a randomized block design in which block size was randomly allocated between 2 and 4 to ensure that the size of the intervention and control groups was equivalent. […] Children randomized to the control group received routine care […]; children who were randomized to the intervention group received an asthma education and management intervention.
Judgment: HIGH
Rationale: Given that this was an educational intervention, this was an unblinded trial. Using blocks of 2-4 would allow for easy determination of the group allocation by the investigators prior to assignment.
Citation: Farber HJ, Oliveria L. Trial of an asthma education program in an inner-city pediatric emergency department. Pediatric Asthma, Allergy & Immunology 2004;17:107-115.
[back to top] [Other sources of bias]